Research On Insulator Coating Defect Detection System Based On Image Processing

Insulator is a disk insulator hung on the connection tower of high voltage wire,which increases the creepage distance,and makes the live parts insulated and plays a role of mechanical support and positioning.But when insulators are exposed to the open air for a long time,the dust in the atmosphere will precipitate to the surface and form dirty layer to produce pollution flashover.In order to prevent the phenomenon of pollution flashover,a protective coating will be plated on the surface of the insulator.However,insulators are exposed to outdoor for a long time.It is necessary to detect insulators regularly and check whether there are defects in the coating.The original method is based on artificial detection of insulator defects,and the detection efficiency is low and prone to leak detection.The application of image processing technology to the detection of insulators can not only reduce the labor cost,but also improve the detection efficiency and detection precision.RTV coating defect detection system designed in this paper is mainly composed of four rotor UAV and image processing system is composed of two parts: four rotor UAV through insulator image acquisition and storage with telephoto camera;image processing system the first pretreatment process of UAV image collected,then a complete picture of the insulator obtained by image finally,through various splicing algorithm show the defect detection of the insulator through GUI interface.The main work is as follows:(1)image acquisition: a four rotor unmanned aerial vehicle with a long focus camera is used to collect the image remotely through the console.(2)image preprocessing: because of the problems of noise and background diversification during the process of image acquisition,we need to preprocess the collected pictures.It mainly includes image grayscale processing,image enhancement technology and image denoising.In this paper,the histogram equalization and median filter algorithm are used to preprocess the image to enhance the image quality and eliminate the image noise.(3)image mosaic: in the process of image acquisition,the target images will be photographed at multiple angles,and the images of two overlapped positions can be stitching,so that the defective parts of insulator can be integrated into a picture now.The key technology of image mosaic is image registration,and the SIFT image registration algorithm is the main application of the defect detection system in this paper.(4)defect detection: when the insulator coating defect texture clear,mainly use the Canny operator to detect defects,if the defect of coating texture is relatively shallow,mainly through three kinds of defect detection algorithms,mainly for the Gabor transform method and gray level co-occurrence matrix,Fu Liye transform.shows the result: the final written in MATLAB2008 a GUI user interface,to detect the defects result through the GUI interface display.